HBase vs PostgreSQL?

I am creating a flink application in Hadoop cluster. I need a datastore for this application.

Operations on datastore:

1. Simple table with 3-4 columns. Equal reads and writes (mostly update)

   • Max - 1000 read/writes per second
   • Average - 100 read/writes per second

2. Table with 10-15 columns. Mostly write and very few reads.

   • Max - 500 read/writes per second
   • Average - 50 read/writes per second

Requirements:

• I will not perform any aggregation like average, max or more complicated operations.

• Most of the reads are on primary keys.

Available options for me:

• HDFS - This is for batch processing. So not fit for my use case.
• Hive - Again more towards batch processing. But I can perform HQL.
• HBase - It is distributed. Good for reads on row keys. No issue with large growing data.
• PostgreSQL(RDBMS) - As number of operations per second is not much, RDBMS can also be good . Easy to use SQL query. Good ORMs in market.

Correct me if I am wrong about these.

I am confused between HBase and PostgeSQL. What would be the good option for this usecase?